Vote Report PH: The Gist So Far in Mapping the 2010 Philippine Elections

05/10/10 - 6:14 PM

We are now on the 11th hour of monitoring in the first ever Philippine automated elections using the Ushahidi-based system VoteReportPH. The Philippine edition seeks to apply lessons learned from past similar implementations, having already received over 200 verified and well-documented reports.

How it Works

The VoteReportPH team has an extensive nation-wide network of poll watchers and monitoring volunteers covering all provinces, reaching almost 80% of all municipalities. Poll watchers send reports via SMS to an online information system where a 24-hour monitoring team collates and verifies incoming information.

These verified reports are then published in the website of Kontra Daya, an election watchdog working parallel with VoteReportPH, of which our Ushahidi implementation keeps track of via feeds. This SMS-to-Feed approach was adopted to ensure only accurate reports and no duplicates were reflected on the map. Reports could also still be made through a web form and the universal election twitter hash tag #juanvote and #votereporph.

Monitors moderated all incoming media, prioritizing the feed reports. Web form and twitter reports are then accepted but left unverified until it is confirmed that they have no duplicate reports among the Kontra Daya feed items. Alerts and updates are then sent en-masse to all VoteReportPH municipal coordinators by the team via SMS blast tool Frontline SMS.

Initial Experiences

VoteReportPH seeks not only to provide a visualization of large-scale fraud and automation glitches, but also to serve as a tool to monitor events that can be responded to by our volunteers. By keeping only the unique, high-quality reports, VoteReportPH’s reports can serve as basis for prompt protest, legal or technical actions.

The first five hours of country’s first automated national elections were characterized by delayed voting, a slow and haphazardly organized voting system, and various PCOS machine malfunctions. There were also sporadic reports of cases of poll-related violence in several areas. An early VoteReportPH alert in Laguna on large-scale fraud and electoral violence initially went under the radar but were promptly reported to the media and responded to by volunteers.

The hash tag #juanvote was popularized for all election-related information. It became hard to monitor it for verification even with numerous volunteers on deck, what with thousands already flooding into the system. Web forms were the least utilized, with voters and monitors staying on the ground with no access to computers. Reports came primarily through SMS.

Initial Conclusions

The VoteReportPH experience had an output of high quality reports that were actionable. Capturing the gravity of fraud on the nation-wide scale however remains a limited effort. Crowdsourcing, after all, is about utilizing a decentralized array of volunteers to accomplish a complex task. In-depth reports from poll watchers around the nation were collated and visualized in real-time.

Twitter reporters could have utilized a trending topic specified on electoral fraud and monitoring. #JuanVote, after all, is a citizen journalism effort that also involved voter education and political analysis. While aggregating the reports is a snap, sifting through the information requires time and effort.

We will see how things go as we near the end of the voting period. We still have to monitor the electronic transmission of votes and canvassing. The data analysis must come as the reporting flows. Here’s to hoping that our technical and political work pays off in not only ensuring the integrity of our democracy, but also in contributing to the advancement of New Politics in the Philippines. # # #